Strange issue with mean, max, and min of vector?

Hello,

I am having an issue with finding the mean max and min of a vector (attached). All values in the vector are positive (increments of time) and less than 1. However, when I test the mean() = -8.0219, max = 9.5770, min = -4.3200e+04. When I try to find() the indices of the min and max, it says that it is 1 but, the value at index 1 is 0.012.

When you test the vector, do you get the same result? Why is it outputting these incorrect values?

Answer 1

[Star Strider]

See the plotted values —

load('diffT.mat')
whos('-file','diffT')

  Name          Size            Bytes  Class     Attributes

  diffT      5370x1             42960  double              

% diffT

neg_vals = nnz(diffT < 0)

neg_vals = 1

[lo,hi] = bounds(diffT)

lo = -4.3200e+04

hi = 9.5770

diffT_median = median(diffT)

diffT_median = 0.0170

diffT_mean = mean(diffT)

diffT_mean = -8.0219

figure
plot(diffT)
grid
xlabel('Index')
ylabel('diffT (Everything)')

figure
plot(diffT)
grid
xlabel('Index')
ylabel('diffT (Bounded)')
ylim([min(diffT(1:1000))  max(diffT(1:2000))])

diffT_out = filloutliers(diffT,'nearest');                  % Detect & Fill The Outlier
[lo,hi] = bounds(diffT_out)

lo = 0.0130

hi = 0.0210

median_diffT_out = median(diffT_out)

median_diffT_out = 0.0170

mean_diffT_out = mean(diffT_out)

mean_diffT_out = 0.0165

stdev_diffT_out = std(diffT_out)

stdev_diffT_out = 7.1457e-04

figure
plot(diffT_out)
grid
xlabel('Index')
ylabel('diffT\_out')

figure
histfit(diffT_out, 100, 'lognormal')
grid

Your statement ‘All values in the vector are positive’ seems not to be the actual situation. There is one value that is equal to -43200, and that sort of spoils the party. You might want to see how that large negative value managed to get in your ‘diffT’ data.

When in doubt, plot the data!

EDIT — Corrected typographical errors.

.

[FsC]

Thank you for the thorough answer! Now I know to plot in the future

[Star Strider]

As always, my pleasure!